Determining the gender of American martens (Martes americana) and fishers (M. pennanti) from track plate stations would significantly augment the information currently gathered from this simple and inexpensive survey method. We used track-plate impressions collected from captured individual martens and fishers of known gender to develop methods to distinguish the gender from tracks. We collected marten tracks from two subspecies M. a. sierrae (tracks n = 54, individuals m = 12, f = 8) and M. a. humboldtensis (tracks n = 109, individuals m = 10, f = 9). We developed a model for martens which performed extremely well (91–100%) using only total track length. For fishers, we collected 64 tracks (from 7 males, 17 females) and developed a 3-variable model, using interdigital pad height, interdigital pad width, and total track length, which correctly classified 98.6% of the tracks. We evaluated track quality on 129 sheets of contact paper collected during summer (dry-season) and 137 sheets collected during fall (wet-season); 100% of dry season compared to 51% of wet season sheets had suitable quality tracks. Gender discrimination of Martes tracks can be used to investigate gender differences in distribution and habitat (associations), but the generality of these findings to other populations is undetermined.